36

CHAPTER III

RESEARCH METHOD

This chapter discusses about sources of data, research design,

research setting, population and sample of research, variables and

indicators of research, methods of data collection, and methods of data

analysis.

A. Research Design

This research is quantitative because the result of the

students’ achievement of pre-test and post-test will be explained

by number, which are evaluated and also interpreted by statistic

procedure. In this case refers to use T-test.

Research method used here is quasi-experiment, a part of

experimental research that do not use control group or it can

called by one group pretest-posttest design1. In this research the

researcher only use one group as experimental group and to get

the value of independent variable is from pre-test score, so post-

test score will be the dependent variable after the researcher

giving the treatment. This research design will be described as

follow:

1 William R. Shadish and Thomas D. Cook, Experimental and

Quasi-Experimental Designs for Generalized Causal Inference, (USA:

Houghton Mifflin Company, 2002), p. 108

E 01 X 02

37

Where,

E : Experimental group/class

X : Treatment

01 : Pre-test

02 : Post-test

B. Research Setting

1. Time of the research

This research will be conducted from February 1st to

February 23rd

2016 counted since the proposal was submitted

until the end of research.

2. Place of the research

This research will be conducted at MA At-Tawaazun,

located in Jl. Nuri no. 56, kel. Widuri, kab.Pemalang.

C. Subject of The Research

1. Population

Population is all of the research of subject.2The

population in this research is all students of grade eleventh

MA At-Tawaazun in academic year of 2015/2016. But it is a

new school, so the number of the entire students is 22

students.

2Suharsimi Arikunto, Prosedur Penelitian Suatu Pendekatan

Praktik, p. 173

38

2. Sample and Sampling

Sample is some of chosen population using certain

procedure and has same characteristic so that can be expected

to represent its population.3 Sampling is the process done to

choose and take sample correctly from population so that it

can be used as valid representative to the population.

Based on the situation and condition of that school,

the researcher decides to take boring sampling or total

sampling. Boring/total sampling is the sample that represents

its population, and it usually taken because the population has

the small number less than 100.4

D. Variable of Research

According to Fred D. Kerlinger as cited by Arikunto, that

all experiments have one fundamental idea behind them; to test

the effect of one or more independent variables on a dependent

variable (it is possible to have more than one dependent variable

in experiments)5, and also quasi experimental.

This research use two variables, independent and

dependent variables. According to Robert as cited by Noor,

3Nana Sudjana and Ibrahim, Penelitian dan Penilaian Pendidikan,

(Bandung: Sinar Baru Bandung: 1989), p. 85.

4Juliansyah Noor, Metodologi Penelitian Skripsi, Tesis, Disertasi,

dan Karya Ilmiah. edisi pertama, (Jakarta: Kencana Prenada Media Group:

2011), p.156

5Suharsimi Arikunto, Prosedur Penelitian Suatu Pendekatan

Praktik, P.159

39

independent variable is a cause that predicted by some of changes

in dependent variable. And dependent variable is main factor that

will be explained, predicted, and influenced by other factors.6

E. Technique of Data Collection

1. Test

Test is any procedure for measuring ability,

knowledge, or performance.7 In order to discover by questions

or practical activities, what someone knows, or what someone

or something can do or is like.

In tests, researchers have at their disposal a powerful

method of data collection, an impressive array of tests for

gathering data of a numerical rather than verbal kind.8 The

data was collected by giving written test. Written test was

conducted twice, there are pre-test and post-test.

This technique is applied by researcher to know the

students’ achievements that have done in learning English,

especially in teaching narrative text on the reading class.

6Juliansyah Noor, Metodologi Penelitian Skripsi, Tesis, Disertasi,

dan Karya Ilmiah. edisi pertama, p. 48-49

7Jack C. Richard. Longman Dictionary of Language Teaching and

Applied Linguistic 2nd edition, p. 546

8Louis Cohen, Lawrence Manion and Keith Morrison, Research

Method in Education. Fifth Ed, (London: Routledge Falmer: 2005), p.317.

40

2. Documentation

Another data is needed to help the researcher run the

research. In addition to do that, data will be collected through

documentation of the students' previous examination score

from the school. It will be used to validate the sample. And

documentation of students’ written test recording is used to

evaluate students’ reading skill.

F. Method of Data Analysis

1. First Phase (Try-out Test)

Try out test is necessary since the result will be used

to make sure that the measuring instrument has such

characteristics as validity and reliability. The instrument to be

tried out was the composition test. The result of test was used

to find out the validity and reliability.

a. Validity

Heaton states that validity is the extent to which

it measures what is supposed to measure and nothing

else.9 The result was consulted to critical score for r-

product moment. If the obtained coefficient of correlation

was higher than the critical score for r-product moment, it

meant that a paragraph was valid at 5% alpha level

9J. B. Heaton, Writing English Language Test, p. 153.

41

significance. To calculate the validity, the researcher used

the formula10

as follows;

∑ (∑ (∑

√{ ∑ (∑ }{ ∑ (∑ }

Where:

= coefficient of correlation between X and Y

N = the number of students

∑X = total score of test item

∑Y = total score

∑XY = the sum of multiplication X times Y

b. Reliability

Reliability refers to the stability or the

consistency of the test scores. Heaton states that

reliability is a necessary characteristic of any good test;

for it to be valid at all, a test must first be reliable as a

measuring instrument.11

In this study, the reliability of

the test was measured by comparing the obtained score

with r-score product moment. Thus, if the obtained score

was higher than the table r-score, it could be said that the

test was reliable.

To calculate the reliability of the test, the

researcher used half split formula with r-score product

10

Suharsimi Arikunto, Dasar-dasar Evaluasi Pendidikan, (Jakarta:

Bumi Aksara: 2007), p. 72-75.

11J. B. Heaton, Writing English Language Test, p. 155

42

moment formula and continues it with spearman-brown

formula as follows12

,

⁄

⁄

( ⁄

⁄

Where:

= index reliability

⁄

⁄ = index correlation between half split data

Next step is consulting the result score of with

score of r . If > test item is reliable.

c. Index of Difficulty

Heaton states that “the index of difficulty of an

item simply shows how easy or difficult the particular

item proved in the test”.13

If a teacher knows deeply about

item difficulty in making a test, he can make his test

easy, medium, or difficult.

To know the item difficulty, the writer used the

formula:

Where:

P = index of difficulty

B = the number of students who answers correctly

12

Suharsimi Arikunto, Prosedur Penelitian Suatu Pendekatan

Praktik, p. 223-224.

13J. B. Heaton, Writing English Language Test, p.172.

43

JS = the total number of students

The index of difficulty level can be classified as

follows:

0.00 < 0.30 is difficult

0.30 < 0.70 is medium

0.70 < 1.00 is easy

d. Discriminating Power

Item of discrimination power tells how well the

item performs in separating the better students from the

poorer students. If the good students tend to do well on

an item and the poor students do badly on the same item,

then the item is a good one because it distinguishes the

good student from the bad student. Heaton states, “The

discrimination index of an item indicated the extent to

which the item discriminated between the testee,

separating the more able testee from the less able. The

index of discriminating power told us if students who

perform well on the whole test tended to do well or badly

on each item in the test.”14

To calculate the index of discriminating power,

the writer used the formula:

14

J. B. Heaton, Writing English Language Test, p.173.

44

Where:

= Number of all students in the upper group

= Number of all students in the lower group

= Number of students in the upper group who

answered the item correctly

= Number of students in the lower group who

answered the item correctly

= The proportion of the upper group who

answered the item correctly

= The proportion of the lower group who

answered the item correctly

The criteria of determining the index of discriminating

are below:

D = 0.00 – 0.20: Poor

D = 0.21 – 0.40: Satisfactory

D = 0.41 – 0.70: Good

D = 0.71 – 1.00: Excellent

2. Second Phase

a. Normality test

The researcher did normality test in this research

to know the normality distribution of class or not. The

normality test with Liliefors is done to find out

distribution data. The steps of normality test use Liliefors

test:

45

1) Hypothesis

Ho : the sample was from population which

normal distributed.

Ha : the sample was not from population which

normal distributed.

2) Statistical Test

| ( ( |

Where

Explanation:

zi = standard value (i)

xi = sample of data (i)

( (

(

3) Level significant (α) = 0,05

4) Critical Area (CA) = { | } with n is size of

sample.

5) Test decision

Ho rejected if Lo in critical area.

6) Conclusion

a) Sample was from population that normal

distributed, if Ho accepted.

b) Sample was not from population that normal

distributed, if Ho rejected.

46

3. Final Phase

Post-test is held after all treatments are conducted.

This test was used to measure student’s achievements after

they were given treatments. The result of test was analyzed

statistically by looking for the mean and the percentage of the

result. The aim is to compare between the pre-test and post-

test. Then, the overall result is counted by using t-test formula

in order to know the significances of the research. The

formula that is used in the paired t-test as follows:15

√

(

Where:

Md = Deviation mean between pre-test and pos-test

= Deviation difference from deviation mean

∑ = Amount of deviation quadrate

N = Number of subject

= or Db is N-1

Testing criteria that apply Ha is accepted if

with determinate ( and the significantα

( α .

15

Suharsimi Arikunto, Prosedur Penelitian Suatu Pendekatan

Praktik, p. 349-352.